Nanoscale

Published on 05 November 2013. Downloaded by University of Reading on 25/01/2018 22:35:08.

PAPER

View Article Online View Journal | View Issue

Cite this: Nanoscale, 2014, 6, 962

Nanoscale mechanics by tomographic contact resonance atomic force microscopy† Gheorghe Stan,*ab Santiago D. Solares,b Bede Pittenger,c Natalia Erinac and Chanmin Suc We report on quantifiable depth-dependent contact resonance AFM (CR-AFM) measurements over polystyrene–polypropylene (PS–PP) blends to detail surface and sub-surface features in terms of elastic modulus and mechanical dissipation. The depth-dependences of the measured parameters were analyzed to generate cross-sectional images of tomographic reconstructions. Through a suitable normalization of the measured contact stiffness and indentation depth, the depth-dependence of the contact stiffness was analyzed by linear fits to obtain the elastic moduli of the materials probed. Besides

Received 17th September 2013 Accepted 1st November 2013

elastic moduli, the contributions of adhesive forces (short-range versus long-range) to contact on each

DOI: 10.1039/c3nr04981g

material were determined without a priori assumptions. The adhesion analysis was complemented by an unambiguous identification of distinct viscous responses during adhesion and in-contact deformation

www.rsc.org/nanoscale

from the dissipated power during indentation.

I.

Introduction

Atomic force microscopy (AFM) has proven to be the most versatile tool in probing mechanical properties of nanoscale volumes.1–7 Various AFM capabilities in forms of spectroscopic measurements and two-dimensional scans have been developed to characterize elastic,8–10 viscoelastic,11,12,14 and adhesive15,16 properties of materials at the nanoscale. However, fully threedimensional (3D) characterization of the tip–sample interactions17–22 has been previously used mostly at the atomistic scale to detail the force and energy elds above atomically at surfaces such as graphite,19 mica20 or ionic surfaces.21,22 In these studies, frequency modulated AFM measurements were performed either in a series of planes at various heights above the investigated surface, followed by a subsequent reconstruction to include dri corrections19,21 or by recording the z-modulation along each scan line.20,22 The tip–sample interactions, probed either in an ultrahigh vacuum17–19,21 or liquid,20,22 were essentially non-contact interactions, with frequency changes in the range of Hz to tens of Hz. Only few attempts were made to extend the 3D analysis for tip–sample contact interactions.23,24 Applications of such nanoscale tomographic characterization can be envisioned for high spatial resolution and in-depth quantitative measurements of mechanical properties of nanocomposites25 and biomaterials,26,27 a

Material Measurement Laboratory, National Institute of Standards and Technology, Gaithersburg, Maryland, 20899, USA. E-mail: [email protected]

b

Department of Mechanical Engineering, University of Maryland, College Park, Maryland, 20742, USA

c

AFM Business Unit, Bruker-Nano Inc., Santa Barbara, California, 93117, USA

† Electronic supplementary 10.1039/c3nr04981g

information

962 | Nanoscale, 2014, 6, 962–969

(ESI)

available.

See

DOI:

in which cases identication and quantitative characterization of subsurface structural modications can be used to understand the nanoscale structure–function relationship of materials. In this work we demonstrated a room-temperature 3D frequency modulated in-air AFM technique to probe the intimate contact mechanical response of a non-at elastomeric surface. The frequency changes associated with the tip–sample interactions were in the range of tens of kHz and no dri corrections were considered for the nanoscale spatial resolution of this investigation. At each position in the scan, force–distance AFM curves were performed with a small sinusoidal modulation applied to one of the eigenmodes of the AFM cantilever.28

II.

Experimental details

The polymers used in this study were polystyrene (Mn ¼ 70 400, Mw ¼ 73 000) from Polymer Sources, and isotactic polypropylene (Mw z 250 000, average Mn z 67 000) from SigmaAldrich. A blend of PS and PP in ratio 70 : 30 wt% was prepared in xylene solution and spin-coated on the Si substrate at a 2000 rpm rate. The obtained lm (about 300 nm thick) was in the form of an amorphous PS matrix with PP regions exhibiting a semi-crystalline structure. The frequency modulation that we used to interrogate the mechanical response of these materials was in the form of a small-amplitude modulation that allowed the observation of the change in resonance frequency as the tip was brought in and out of contact with a material. As shown in Fig. 1, when the tip contacts a material, the boundary conditions of the cantilever– tip system change and the eigenmodes of the out-of-contact cantilever shi accordingly with the change in tip–sample

This journal is © The Royal Society of Chemistry 2014

View Article Online

Published on 05 November 2013. Downloaded by University of Reading on 25/01/2018 22:35:08.

Paper

Nanoscale

of 150 Hz made by the PLL in determining the resonance frequency shi when the photodetector signal is perfectly locked at the target phase and the resonance peak experiences a drop to one h in its effective quality factor due to dissipation. Such attenuations in the quality factor were observed routinely in our measurements in the case of the largest measured frequency shis. These errors have negligible contributions to the calculated elastic moduli (from the measured resonance frequencies) and dissipated power (from the measured resonance frequencies and amplitudes); for both measured frequency and amplitude the errors are proportional to the frequency shis. Based on these considerations, no corrections were applied in the present work to the measured PLL resonance frequencies and amplitudes.

When an AFM probe is brought into contact with a material, the cantilever deflects back due to an increase in contact stiffness. This also changes the dynamics of the cantilever as the resonance frequency of the cantilever is sensitive to the stiffness of the cantilever–tip–sample system. The tip–sample contact was considered here as represented by a spring coupled in parallel with a dashpot to account for the conservative and dissipative responses of the contact. Fig. 1

contact stiffness.29,30 The resonance frequency changes were tracked through a phase-locked loop (PLL) detection scheme in which the cantilever’s base was driven by a constant-amplitude excitation and the excitation phase was adjusted to lock at the value selected during tuning of the cantilever, close to 90 degrees (this method is referred to as a constant-excitation PLL31,32). We used this to simultaneously observe the change in the resonance frequency and amplitude of the second eigenmode of the cantilever over an area comprising a few PP spherulite-like regions surrounded by large PS areas. A LabView (National Instruments, Austin, TX, USA) controlled module was interfaced between a MultiMode 8 Bruker AFM (Bruker, Santa Barbara, CA, USA) and a Nanonis PLL (SPECS Zurich GmbH, Zurich, Switzerland) for data acquisition. The AFM was operated in force volume mode, with piezo ramps of 75 nm and a contact peak force of 25 nN. The thermal noise spectrum analysis was used to calibrate the rst two vibration eigenmodes of the PPP SEIH cantilever used (Nanosensors, Neuchatel, Switzerland) in terms of resonance frequency, spring constant, and Q-factor: f1air ¼ 104.9  0.1 kHz, 1 air kair c1 ¼ 9.5  0.5 N m , and Q1 ¼ 265  10 for the rst mode and air air air air 2 1 f2 ¼ 657.6  0.1 kHz, kc2 ¼ kair c1 ( f2 /f1 ) ¼ 373.3 N m , and air Q2 ¼ 770  20 for the second mode. The ramps were distributed on a 128  128 grid over a 2 mm  2 mm area and were performed at a constant frequency of 1 Hz (1 second per ramp). Four signals (AFM diode, AFM piezo, PLL frequency, and PLL amplitude) were acquired simultaneously at a rate of 1 kHz. The PLL was operated in constant excitation mode with a driving excitation of 500 mV added to the piezo shaker of the AFM cantilever. The frequency shi was tracked within a locked window of 39.1 kHz around the second free-resonance frequency and no dris were observed outside of this window. Using harmonic oscillator theory we calculated an error of order

This journal is © The Royal Society of Chemistry 2014

III. Tomographic measurements and visualization Maps of each measured quantity can be reconstructed at any indentation depth using data from either approach or retract. In Fig. 2 are shown a few resonance frequency and resonance amplitude measurements obtained during approach, aligned with the “coming into contact” direction of the approaching AFM tip. In each panel are shown constant indentation-depth maps (horizontal cross-sections of the tomographic 3D data) for

Fig. 2 Resonance frequency (left panel) and amplitude (right panel) of the second eigenmode during approach at various tip–sample indentation depths over a 2.0 mm  0.7 mm area comprising PP and PS regions. The resonance frequency and amplitude contrast between PP and PS regions changes with the tip–sample separation. The color scale was adjusted in each image to enhance the contrast.

Nanoscale, 2014, 6, 962–969 | 963

View Article Online

Published on 05 November 2013. Downloaded by University of Reading on 25/01/2018 22:35:08.

Nanoscale

resonance frequency (le panel) and amplitude (right panel), going from d ¼ 5 nm (out of contact) to d ¼ 5 nm (in contact). The PP and PS regions exhibit different contrast in both the frequency and amplitude maps. The PP regions (the large circular region on the right side of the maps and few small dots) remain the only ones visible in the maps at indentations larger than the maximum indentation depth on PS (about 2 nm). As the tip was pushed into contact, the resonance frequency continuously increased from 657 kHz (free resonance) to about 690 kHz (deepest indentation). The frequency contrast between the two materials starts to become visible in the adhesive range. Thus, at d ¼ 2.5 nm, a positive shi in the resonance frequency is observed over the PP region whereas the PS region is still at the free resonance frequency, which means that no signicant interaction between the tip and PS was established at this distance. With the further reduction of the distance between the tip and the surface, a frequency response is observed for both materials: at d ¼ 1.0 nm, the resonance frequency is about 664 kHz on PP and about 660 kHz on PS. Thus, the resonance frequency on PP is larger than on PS, which suggests an increased stiffness response of PP compared with PS over the adhesive range. The frequency contrast between PS and PP almost vanishes around the zero contact point. It is worth noting here that the resonance frequency (and stiffness) response, especially in the adhesive range, does not reproduce directly the elastic contrast between the two materials as the contact stiffness depends on the elastic modulus and contact area. The contact area in turn depends on the applied force, adhesive force, tip geometry, and elastic modulus. It is thus necessary to consider contact mechanics in correlating resonance frequencies (and stiffnesses) with the elastic moduli of materials. Beyond the contact point, at positive indentation depths, a clear distinction in the contact resonance frequency is observed as the tip further indents into the materials. Along the common indentation range for both materials (d between 0 and 2 nm), the contact resonance frequency is larger on PS than on PP, which means that PS is stiffer than PP. Because of the stiffness difference, deeper indentations were experienced on PP than on PS for the same applied peak force of 25 nN. This is why, as can be seen in Fig. 2, at depths greater than 2 nm, the measured resonance frequency and amplitude were available only over PP regions. In the amplitude panel, it can be observed that, as the tip was brought into contact, the resonance amplitude was larger on PS than on PP at any indentation depth. In addition to a discrete sequence of constant-depth maps like the one shown in Fig. 2, an even more complete visualization can be obtained in tomographic cross-sections of the volume of measurements based on the known 3D space location of each measurement. Figs. 3(B)–(E) show such tomographic cross-sections of contact resonance frequencies and amplitudes from data collected during measurements over the area shown in Fig. 3(A). Each image comprises three cross-sections: the xz and yz planes, perpendicular to the sample’s surface, are sections across PS/PP regions and along a PP domain, respectively, and the xy plane is a cross-section parallel to the sample surface at zero indentation depth. The images shown in Fig. 2 and 3 were extracted from the same set of measurements.

964 | Nanoscale, 2014, 6, 962–969

Paper

Tomographic cross-sectional maps of CR-AFM measurements. (A) AFM micrograph showing the surface topography of a PS–PP blend polymer. (B and D) Tomographic cross-sectional images of contact resonance frequency and (C and E) amplitude of the second eigenmode of the cantilever over the area shown in (A). The vertical xz and yz crosssections of the tomographic images are along the cross-lines shown in (A). The images in (B) and (C) were reconstructed from data collected during probe approaches and (D) and (E) from data collected during retracts. The PP and PS regions are visually delimited in the horizontal tomographic planes by their contrast in frequency and amplitude.

Fig. 3

In contrast with a two-dimensional CR-AFM mapping performed at a given applied force,10,33 tomographic CR-AFM measurements provide additional information in the form of the depth-dependences for both contact stiffness and dissipation that can be analyzed in great detail at any point in the scan. Besides the in-depth contact analysis, the 3D measurement provides also insight into the formation and breaking of the contact between the tip and the material probed. In our case,

This journal is © The Royal Society of Chemistry 2014

View Article Online

Paper

the formation and breaking of the contact can be seen to differ for each material and also from one material to the other. This is clearly illustrated by the color contrast of the horizontal planes of Fig. 3(C) and Fig. 3(E) between the PS and PP regions.

Published on 05 November 2013. Downloaded by University of Reading on 25/01/2018 22:35:08.

IV. Results and discussion The measured resonance frequencies were rst converted into contact stiffness by using a spring-coupled-clamped beam model for the mechanically vibrated cantilever.29 Then a contact mechanics model describing a dynamically indented elastomer by a at punch was used to t the depth-dependence of the contact stiffness at every location in the scan and extract the elastic modulus and a transition parameter at that location. In addition to the conservative response, the dissipative part of the contact interaction was also analyzed in terms of dissipated power based on the measured resonance frequencies and amplitudes.13,14 We will analyze rst the conservative response and calculate the elastic moduli of the materials probed and then extract the dissipative component from the dynamics of damped eigenmode vibrations. Although the 3D maps shown in Fig. 3 can be directly converted into their corresponding 3D maps of elastic modulus and dissipation, we focused here, for methodological purposes, on analyzing the details of the depthdependence of the contact stiffness and dissipation. The contact mechanics of adhesive contacts on perfectly elastic materials is customarily analyzed within the limits of two models that include contributions of adhesive forces: Derjaguin, M¨ uller, and Toporov (DMT) model,34 which considers the contribution of long-range attractive forces outside the contact area, and Johnson, Kendall, and Roberts (JKR) model,35 which includes the contribution from attractive forces acting only inside the contact area. The two theoretical models are used accordingly with the measurement conditions: either small radius tips on stiff materials with low surface energies (DMT model) or large radius tips on compliant materials with high surface energies (JKR model); simple analytical expressions can be derived for the depth-dependence of the contact stiffness with any of these models.28,36 A more realistic interpretation relies on using a model that captures the transition regime between these two limiting cases, e.g. Maugis-Dugdale,37 Carpick–Ogletree–Salmeron,38 or Schwarz39 models. Due to its simple analytical equations and direct interpretation, we used in this work the Schwarz model to analyze both the approach and retract portions of the stiffness–force curves. In the Schwarz model, the transition parameter s1 (dened as the square root of the ratio between the work against the short-range adhesive forces and the total work of adhesion) ranges from 0 (DMT limit) to 1 (JKR limit). In our analysis based on the Schwarz model, in order to observe the intrinsic adhesive response of each material, no a priori assumption was made for the value of s1, which was instead considered a free t parameter. In this way, the contribution of long and short range adhesive forces to the formed tip–sample contact can be determined (not assumed) on each material. The adhesive tip–sample contact was modeled as a spherical tip dynamically indenting an elastomer. As

This journal is © The Royal Society of Chemistry 2014

Nanoscale

previously shown40,41 and also observed in our data (positive contact stiffness at the maximum negative applied forces), the stiffness response of a dynamically indented elastomer is equivalent to that of a at punch with the contact radius of the static conguration (Schwarz model in our case). In this dynamic at-punch approximation, the “pinned” behavior of the contact during high-frequency oscillations is determined by the retarded viscoelastic response of the contact edge to the fast motion of the tip. Within the Schwarz model, the indentation depth-dependence d of the contact stiffness k* in the dynamic at-punch approximation can be calculated from (see ESI† for details): pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi   d ¼ k*2 = 4RE *2  h 2k* Fa =ðRE *2 Þ; (1) pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi2 where h ¼ s1 = 4  s1 , E* is the reduced tip–sample elastic modulus expressed in terms of the indentation moduli of the indented sample Ms and the indenting tip MT, E* ¼ (1/Ms + 1/ MT)1, Fa is the maximum adhesive force, and R is the tip radius; for compliant materials, E* is comparable with the indentation modulus of the indented sample Ms as the contribution from the much stiffer tip, 1/MT, is negligible. The indentation modulus M of an isotropic elastic material is expressed in terms of Young’s modulus E and Poisson’s ratio n, M ¼ E/(1  v2). While the rst term in eqn (1) contains the elastic Hertzian dependence, the second term accommodates the adhesive contribution to the contact stiffness. In Fig. 4(A) and (B) are shown individual curves of the contact stiffness as a function of the indentation depth on PP and PS. From these plots it can be observed that PS is stiffer than PP over the explored indentation range, except for the region around contact formation where meniscus forces enhance the adhesion interaction between the tip and the sample and make PP stiffer than PS. A clear distinction between the two materials was observed in the measured d vs. k* curves as shown in Fig. 4(A). The measurements along with the theoretical curves suggest that for a given tip radius and adhesive force, E* and s1 could be determined as t parameters from the above non-linear equation. However, as follows from eqn (1), the two parameters can ffi be further separated through the linear pffiffiffiffi dependence of d= k* vs. k*3/2, pffiffiffiffiffi d= k* ¼ a þ bk*3=2 :

(2) pffiffiffiffiffiffiffiffi Here, E* is given solely by the slope of eqn (2) as E ¼ 1= 4Rb pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi and s1 by the slope and y-intercept as s1 ¼ 2|a|= 8Fa b þ a2 . This linearization is shown explicitly in Fig. 4(B) for the same curves plotted in Fig. 4(A). Also, as shown in the ESI,† the linearization justies in this case the use of the dynamic atpunch approximation over that of a spherical indenter. Thus, for each material, a distinct slope of the linear region and different ranges of values for s1 (around 0.5 for PS and 1 for PP) can be observed. The analysis was performed at each pixel in the scan to construct the maps of E* and s1 shown in Fig. 5(A) and (B). In Fig. 5(A), the contrast corresponds quantitatively to indentation moduli of 3.2  0.4 GPa for PS and 1.4  0.2 GPa for PP, values comparable with the known elastic moduli of these materials (3.2 GPa to 3.4 GPa for PS and 1.4 GPa to 2.4 GPa for *

Nanoscale, 2014, 6, 962–969 | 965

View Article Online

Published on 05 November 2013. Downloaded by University of Reading on 25/01/2018 22:35:08.

Nanoscale

Paper

Fig. 5 Elastic modulus and transition parameter maps. (A) Map of the calculated indentation modulus (from the slope of the linear part of pffiffiffiffiffi d= k* vs. k*3/2 curves, as shown in Fig. 4(B)) over the entire scanned area. ffi Map of the transitional parameter s1 (from the y-intercept of pffiffiffiffi(B) d= k* vs. k*3/2 curves) over the lower part of the scanned area. (C) Histograms of the calculated values from approach and retract maps over the entire scanned area shown in (B).

Fig. 4 Contact stiffness conversion into indentation modulus. (A) Indi-

vidual curves (continuous line) of measured indentation depth versus contact stiffness on PS and PP; the dashed (PP) and dotted (PS) lines are model curves calculated with eqn (1) for the average values of the involved parameters (R ¼ 15 nm, E* ¼ 1.4 GPa for PP and 3.2 GPa for PS, Fa ¼ 5.5 nN for PP and 4.0 nN for PS) and various values of s1. (B) curves taken from (A) and normalized in accordance with eqn (2) for linear fits.

PP).42 The uncertainties represent one standard deviation of the calculated values. The value of 3.2 GPa for the elastic modulus of PS was considered as a reference and it was obtained for a tip radius of 15 nm. The method of using reference materials is common practice for calibrating CR-AFM measurements;8,10,28,33 the calibration method was used here to infer the tip radius. It is worth noting that the distinct contrast between the elastic moduli of PS and PP obtained here is based on depth-dependence analysis of the contact stiffness whereas little contrast was observed between the elastic moduli of these two materials in CR-AFM mapping at a given applied force.33 As indicated by the individual curves shown in Fig. 4(A) and (B) and revealed by the contrast of the s1 map in Fig. 5(B), a larger contribution from short-range adhesive forces (greater s1 values) was identied to the contact formation on PP than on PS. In the histograms of s1 values (Fig. 5(C)), this contrast is better observed during probe retract (breaking contacts) than approach

966 | Nanoscale, 2014, 6, 962–969

(forming contacts) as enhanced by the difference in contact adhesion hysteresis of these materials. The distinction in the adhesive responses of the two materials can also be rationalized from the depth-dependence of the dissipated power. In Fig. 6(A)–(C) are shown individual curves for the indentation depth dependence of the resonance frequency, amplitude, and dissipated power, respectively, as the probe was brought in and out of contact on PS and PP. In Fig. 6(D)–6(F) (approach), and Fig. 6(G)–(I) (retract), these indentation depthdependences are plotted along a scan line with clear distinction between the PS and PP regions. The dissipated power refers here to the power dissipated by the tip–sample interaction and it was calculated as the difference between the power generated by the cantilever’s driver and the power dissipated by the motion of the cantilever in air.14,32 Along the approach and retract portions of the dissipated power curves the nature of the dissipative processes can be separated and analyzed. At contact formation, comparable maxima in the dissipated power can be observed for each material (red for PP and blue for PS in Fig. 6(C)). The absence of a clear material signature for these maxima suggests that at contact formation the dissipated power is mostly associated with contact-geometry dependent damping due to capillarity and contamination. Once this maximum is passed and the contact is formed, a distinct response can be observed for each material as the tip indents further into the sample: An almost constant dissipated power is observed on PP while a monotonic increasing dissipated power is observed on PS. These portions of the curves (with little hysteresis between approach and retract) could be associated with the intrinsic viscous response of the materials probed. It is conceivable that the reduction in the dissipated power on PP at shallow indentations compared with PS could be associated with the semi-crystalline structure of PP. The damping response of the two materials would eventually become indistinguishable at large applied forces when the contact will behave as a vibration-damping clamp. A characteristic material response can be analyzed also at the break of contact on the retract curves. Large increases in the

This journal is © The Royal Society of Chemistry 2014

View Article Online

Published on 05 November 2013. Downloaded by University of Reading on 25/01/2018 22:35:08.

Paper

Nanoscale

Fig. 6 Depth-dependence of the resonance frequency, amplitude and dissipated power. (A) Contact resonance frequency, (B) amplitude and (C) dissipated power on PS and PP during individual approach and retract spectroscopies. Larger frequencies were consistently observed in the adhesive range on PP than PS, which suggests an enhanced stiffer adhesive response of PP compared with PS in contrast with their contact stiffnesses. While an insignificant hysteresis is observed in frequency, amplitude, and power curves during contact, distinct signatures can be rationalized for PS and PP from their adhesion hysteresis especially in amplitude and dissipated power. (D) to (I), Tomographic cross-sections of contact resonance frequency ((D) approach and (G) retract), amplitude ((E) approach and (H) retract) and dissipated power ((F) approach and (I) retract) from 10 nm to 5 nm indentation depth (bottom is out of contact) along a scan line crossing PS and PP regions. Insight into the distinctive mechanical responses of the two materials can be rationalized from statistics along the scan line for any of the quantities shown on either approach or retract.

dissipated power accompany contact ruptures, over about 4 nm on PS and 6 nm on PP. This also shows that during one approach/retract cycle a much larger amount of power is dissipated during contact formation/breaking than contact deformation itself. Contrary to this, an integral assessment of the dissipated energy per cycle is made in quasi-static AFM force– displacement curve analysis, when the total dissipated energy coming from the difference in work between approach and retract curves is normally associated with the viscoelastic response of a material. However, as the present analysis shows, most of the dissipated energy per cycle is in form of power bursts during contact formation and contact rupture. By comparing the plots and maps in Fig. 6, it appears that the dissipated power behavior follows that of the resonance amplitude in the contact zones but is specically tailored by the resonance frequency in the adhesion zone. This points to the above analysis, with s1 recovered from the adhesion part of This journal is © The Royal Society of Chemistry 2014

the contact stiffness curves. A larger amount of dissipated power characterizes the contact rupture on PP than on PS and this is associated with a greater value of s1 for PP than for PS.

V. Conclusions In this work we demonstrated depth-sensing of contact stiffness and dissipated power measurements that are capable of differentiating the elastic and viscoelastic responses during contact formation/breaking and contact deformation on viscoelastic surfaces. This was achieved through a unique combination of contact resonance and amplitude measurements during regular AFM force volume scans. The ability to measure the change in the dynamics of the cantilever as the AFM probe is brought in and out of contact with the sample during scanning provides a detailed 3D characterization of the mechanical response of the near-surface of the sample. Any necessary analysis in terms of

Nanoscale, 2014, 6, 962–969 | 967

View Article Online

Published on 05 November 2013. Downloaded by University of Reading on 25/01/2018 22:35:08.

Nanoscale

contact stiffness and dissipation can be spatially carried out through tomographic cross-sections of the 3D CR-AFM data. As shown in the current investigation on a PS–PP blend, this type of measurement provides characterization for both the conservative and dissipative near-surface mechanics. We demonstrated the accuracy of the new method by extracting depth-averaged elastic moduli from the depthdependent contact stiffness measurements. The necessary contact mechanics for this analysis was provided by a transitional contact model for a dynamically indented elastomer. As the model adjusts the transition between short and long range adhesive forces, the contribution of adhesive forces to contact on each material was determined without a priori assumptions. Based on the proposed analysis, the mechanical properties of the two materials probed, PS and PP, were qualitatively and quantitatively differentiated in reconstructed maps of elastic modulus and transition parameter. In addition, the adhesion characterization was found to be in very good correlation with the analysis of the dissipated power during formation/breaking of the contact. The dissipated power during tip–sample interaction was determined from tomographic cross-sections of the measured resonance frequency and amplitude. With its 3D detailed visualization and quantitative mechanical property measurement capability, the new proposed technique can be used for 3D investigation of the structure–property relationship of various nanocomposite materials and biological interfaces.

VI. Disclaimer Certain commercial equipment, instruments or materials are identied in this document. Such identication does not imply recommendation or endorsement by the National Institute of Standards and Technology, nor does it imply that the products identied are necessarily the best available for the purpose.

References 1 N. A. Burnham and R. J. Colton, J. Vac. Sci. Technol., A, 1989, 7, 2906. 2 P. Maivald, H. J. Butt, S. A. C. Gould, C. B. Prater, B. Drake, J. A. Gurley, V. B. Elings and P. K. Hansma, Nanotechnology, 1991, 2, 103. 3 S. Magonov, V. B. Elings and M. Whangbo, Surf. Sci., 1997, 375, L385. 4 O. Sahin, S. Magonov, C. Su, C. F. Quate and O. Solgaard, Nat. Nanotechnol., 2007, 2, 507. 5 S. Jesse, S. V. Kalinin, R. Proksch, A. P. Baddorf and B. J. Rodriguez, Nanotechnology, 2007, 18, 435503. 6 B. Pittenger, N. Erina and C. Su, Bruker Application Note AN128, 2010. 7 R. Garcia and E. T. Herruzo, Nat. Nanotechnol., 2012, 7, 217. 8 U. Rabe, S. Amelio, M. Kopycinska, S. Hirsekorn, M. Kempf, M. G¨ oken and W. Arnold, Surf. Interface Anal., 2002, 33, 65. 9 H. Wagner, D. Bedorf, S. K¨ uchemann, M. Schwabe, B. Zhang, W. Arnold and K. Samwer, Nat. Mater., 2011, 10, 439.

968 | Nanoscale, 2014, 6, 962–969

Paper

10 G. Stan, S. W. King and R. F. Cook, Nanotechnology, 2012, 23, 215703. 11 B. Gotsmann, C. Seidel, B. Anczykowski and H. Fuchs, Phys. Rev. B: Condens. Matter Mater. Phys., 1999, 60, 11051. 12 P. A. Yuya, D. C. Hurley and J. A. Turner, J. Appl. Phys., 2008, 104, 074916. 13 R. Garcia, C. J. Gomez, N. F. Martinez, S. Patil, C. Dietz and R. Magerle, Phys. Rev. Lett., 2006, 97, 016103. 14 G. Chawla and S. D. Solares, Appl. Phys. Lett., 2011, 99, 074103. 15 A. Raman, S. Trigueros, A. Cartagena, A. P. Z. Stevenson, M. Susilo, E. Nauman and S. Antoranz Contera, Nat. Nanotechnol., 2011, 6, 809. 16 N. Gavara and R. S. Chadwick, Nat. Nanotechnol., 2012, 7, 733. 17 H. H¨ olscher, S. M. Langkat, A. Schwarz and R. Wiesendanger, Appl. Phys. Lett., 2002, 81, 4428. 18 L. Gross, F. Mohn, N. Moll, P. Liljeroth and G. Meyer, Science, 2009, 325, 1110. 19 B. J. Albers, T. C. Schwendemann, M. C. Baykara, N. Pilet, M. Liebmann, E. I. Altman and U. D. Schwarz, Nat. Nanotechnol., 2009, 4, 307. 20 T. Fukuma, Y. Ueda, S. Yoshioka and H. Asakawa, Phys. Rev. Lett., 2010, 104, 016101. 21 S. Fremy, S. Kawai, R. Pawlak, T. Glatzel, A. Baratoff and E. Meyer, Nanotechnology, 2012, 23, 055401. 22 E. T. Herruzo, H. Asakawa, T. Fukuma and R. Garcia, Nanoscale, 2013, 5, 2678. 23 J. E. Schmutz, H. H¨ olscher, D. Ebeling, M. M. Sch¨ afer and B. Anczykowski, Ultramiscroscopy, 2007, 107, 875. 24 E.-C. Spitzner, C. Riesch and R. Magerle, ACS Nano, 2011, 5, 315. 25 L. Tetard, A. Passian and T. Thundat, Nat. Nanotechnol., 2010, 5, 105. 26 M. Stolz, R. Gottardi, R. Raiteri, S. Miot, I. Martin, R. Imer, U. Staufer, A. Raducanu, M. D¨ uggelin, W. Baschong, A. U. Daniels, N. F. Friederich, A. Aszodi and U. Aebi, Nat. Nanotechnol., 2009, 4, 186. 27 A. Fuhrmann, J. R. Staunton, V. Nandakumar, N. Banyai, P. C. W. Davies and R. Ros, Phys. Biol., 2011, 8, 015007. 28 G. Stan, S. W. King and R. F. Cook, J. Mater. Sci., 2009, 24, 2960. 29 U. Rabe, K. Janser and W. Arnold, Rev. Sci. Instrum., 1996, 67, 3281. 30 K. Yamanaka and S. Nakano, Jpn. J. Appl. Phys., Part 1, 1996, 35, 3787. 31 L. Nony and A. Baratoff, Phys. Rev. B: Condens. Matter Mater. Phys., 2006, 74, 235439. 32 H. H¨ olscher, B. Gotsmann and A. Schirmeisen, Phys. Rev. B: Condens. Matter Mater. Phys., 2003, 68, 153401. 33 J. P. Killgore, D. G. Yablon, A. H. Tsou, A. Gannepalli, P. A. Yuya, J. A. Turner, R. Proksch and D. C. Hurley, Langmuir, 2011, 27, 13983. 34 B. V. Derjaguin, V. M. M¨ uller and Y. P. Toporov, J. Colloid Interface Sci., 1975, 53, 314. 35 K. L. Johnson, K. Kendall and A. D. Roberts, Proc. R. Soc. London, Ser. A, 1971, 324, 301.

This journal is © The Royal Society of Chemistry 2014

View Article Online

Paper

40 K. J. Wahl, A. S. Asif, J. A. Greenwood and K. L. Johnson, J. Colloid Interface Sci., 2006, 296, 178. 41 J. A. Greenwood and K. L. Johnson, J. Colloid Interface Sci., 2006, 296, 284. 42 A. Matsuoka and B. Maxwell, J. Polym. Sci., 1958, 32, 131.

Published on 05 November 2013. Downloaded by University of Reading on 25/01/2018 22:35:08.

36 O. Pi´ etrement and M. Tryoyon, J. Colloid Interface Sci., 2000, 226, 166. 37 D. Maugis, J. Colloid Interface Sci., 1992, 150, 243. 38 R. W. Carpick, D. F. Ogletree and M. Salmeron, J. Colloid Interface Sci., 1999, 211, 395. 39 U. D. Schwarz, J. Colloid Interface Sci., 2003, 261, 99.

Nanoscale

This journal is © The Royal Society of Chemistry 2014

Nanoscale, 2014, 6, 962–969 | 969

Nanoscale mechanics by tomographic contact resonance atomic force microscopy.

We report on quantifiable depth-dependent contact resonance AFM (CR-AFM) measurements over polystyrene-polypropylene (PS-PP) blends to detail surface ...
2MB Sizes 0 Downloads 0 Views